This research explores the application of reinforcement learning (RL) algorithms in intelligent manufacturing to improve operational efficiency within simulated environments. By leveraging RL-based optimization strategies, the study aims to enhance decision-making in production scheduling, process control, and resource allocation. The integration of digital twins and high-fidelity simulations allows for iterative learning, enabling adaptive and autonomous system improvements. The proposed approach is expected to reduce production costs, minimize downtime, and optimize performance across manufacturing processes.